Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add alttextai/alttext-mcp --skill alttextgit clone --depth 1 https://github.com/alttextai/alttext-mcpWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/alttextai/alttext-mcp/alttext)<a href="https://agentmods.dev/skills/alttextai/alttext-mcp/alttext"><img src="https://agentmods.dev/badge/skills/alttextai/alttext-mcp/alttext/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/alttextai/alttext-mcp/alttext"><img src="https://agentmods.dev/badge/skills/alttextai/alttext-mcp/alttext.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00093 | $0.01863 |
| Opus 5.5 | $0.00037 | $0.00745 |
| Sonnet 5.5 | $0.00019 | $0.00373 |
| Haiku 4.5 | $0.00009 | $0.00186 |
Grade A, and why
alttext scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 103 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AltText.ai CLI
The alttext command runs every AltText.ai operation from a terminal. Results go to stdout; errors, warnings and notes go to stderr. Each command's flags are listed by alttext <noun> <verb> --help.
Start every session with whoami
Run this before anything else:
alttext whoami --json
It returns account, key_name, key_kind (site or full_access), can_manage_keys and credential_source. Use it to decide what you can do:
- Exit 3, "Not logged in": there is no credential. Tell the person to run
alttext loginthemselves, in their own terminal, and to come back whenalttext whoamiworks. Stop there. can_manage_keys: false: thekeyscommands will fail with exit 4. Key management needs a full-access key, which the person creates in the AltText.ai dashboard and logs in with.can_manage_keys: true: key management works. Read the separation rule below before doing anything that touches untrusted web content.
Never handle the API key yourself
- Never ask for an API key, and never accept one in chat. The person logs in by running
alttext loginin their own terminal, where the key is read from a hidden prompt and stored in the OS keychain. - If the person pastes a key into the conversation anyway, don't use it, don't store it, don't put it in a command or a file, and don't repeat it. Tell them it's now in the chat transcript, that they should rotate it (
alttext keys rotate <id>from a full-access login, or the dashboard), and that they should log in withalttext logininstead. - Never put a key on a command line (
ALTTEXT_API_KEY=... alttext ...) and never read the credentials file or the keychain entry.ALTTEXT_API_KEYis for the person's own CI and shell setup, not for you.
Use --json for anything you read back
Pass --json whenever you parse a result. It prints the same shape the MCP tools return. Human output is for showing the person, not for parsing. Pass --no-input so a command fails instead of waiting on a prompt you can't answer.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- today First seen · 103 lines · 93 tokens per session scan A 9eca41a445ce
alttext is a skill published in the GitHub repository alttextai/alttext-mcp (0 stars, last pushed 2d ago), licensed MIT. It adds 93 tokens to every session and 1,863 once invoked, about $0.0004 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-15.
Other skills, from other repositories
alt-text-writer
Writes descriptive alt text for editorial images following WCAG accessibility guidelines — concise enough for screen readers, specific enough to convey the image's editorial purpose.
chart-description-writer
Writes a precise, accessible text description of a data visualization — covering the chart type, what it shows, the key finding, and the data range — suitable for publication as a caption, alt-text, or standfirst alongside the chart.
antislop-human
Human and accessibility skill for antislop. Contrast, keyboard, focus, and states for real people. Includes the contrast checker.
antislop-layoutmobile
Mobile layout skill for antislop. Use for layouts that reflow across screen sizes, phone to desktop: grids, overflow, tap targets. Load with the core.
data-dashboard
Build a dense, data-heavy screen - an analytics console or a trading terminal - with real charts drawn from tokens: candlestick, volume, depth, stacked area, waterfall, scatter, correlation matrix, donut, gauge, sparklines, heatmaps and order books. Use when the request is a dashboard, terminal, monitoring view, data…
critique
Adversarial design critique of the current work — render it, look at it, and argue for rejection. Run after the gates are green, never instead of them.